Music artist retrieval system and method of retrieving music artist

1. A music artist retrieval system comprising: a music artist similarity map storing section that computes a plurality of similarities for a plurality of music artists, the similarity being computed between one of the plurality of music artists and other music artists based on features of the music artists, and creates a music artist similarity map for the plurality of music artists based on the plurality of similarities thus computed and stores the music artist similarity, the music artist similarity map storing section including: a similarity computing section that computes the similarity between one of the music artists and another one of the music artists based on distances between the plurality of music data for the one music artist and the plurality of music data for the other one music artist; a map creating section that creates the music artist similarity map based on the plurality of similarities for the plurality of music artists computed by the similarity computing section, the map creating section creating the music artist similarity map so that the plurality of similarities computed for the plurality of music artists may have a relationship unicursally drawn in one continuous line according to a traveling salesman algorithm; and a map storing section that stores the music artist similarity map created by the map creating section; a music data storing section that stores a plurality of music data for each of the plurality of music artists; a similar artists selecting and displaying section that, when one music artist is selected from among the plurality of music artists, selects a plurality of music artists whose similarities are close to the similarity of the one music artist, based on the music artist similarity map, and displays on a display a plurality of artist indications corresponding to the one music artist and the plurality of music artists whose similarities are close to the similarity of the one music artist, the similar artists selecting and displaying section including: an initial selecting section that initially selects one predetermined or randomly-selected music artist from among the plurality of music artists at a startup of the system; and a specific artist selecting section that, when one artist indication is selected with an operation of an operator from among the plurality of artist indications displayed on the display after the initial selecting section has initially selected the one music artist, determines one music artist corresponding to the one selected artist indication as the one music artist selected from among the plurality of music artists; a music data playing section that, when one of the plurality of artist indications displayed on the display is selected and then a play command is inputted, extracts at least one music data of the music artist corresponding to the selected artist indication from the plurality of music data stored in the music data storing section, and plays the at least one music data, a term data storing section that stores term data corresponding to the plurality of music artists mapped to the music artist similarity map, the term data being obtained by computing, for each of the plurality of music artists, a term frequency or a term weight defined on the basis of the term frequency in respect of a plurality of predetermined terms in connection with artists and music pieces, picked up from among a plurality of character information data regarding the plurality of music artists; a relevant term display section that displays on the display a term closely associated with the one music artist based on the term data regarding the one music artist stored in the term data storing section in such a manner that the term displayed on the display may be identified as being associated with the one music artist, when the initial selecting section or the specific artist selecting section selects the one music artist; and a pseudo-map graphic display section that displays on the display a closed-loop pseudo-map graphic in which positions corresponding to the plurality of music artists are mapped based on the music artist similarity map, and moves the pseudo-map graphic in conjunction with a movement of the plurality of artist indications displayed by the similar artists selecting and displaying section so that a display position of the artist indication of the one music artist displayed on the display by the similar artists selecting and displaying section may always be consistent with a position of the one music artist mapped to the pseudo-map graphic, wherein the term data storing section stores at least the term data relating to a plurality of terms belonging to a first group, which are considered to be important as a decisive factor in selecting a music artist, and the term data relating to a plurality of terms belonging to a second group, which are considered to be less important than the plurality of terms belonging to the first group but may be usable as a decisive factor; and wherein the relevant term display section displays the top two or more terms having the high term frequency or the high term weight among the plurality of terms belonging to the first group as well as the top two or more terms having the high term frequency or the high term weight among the plurality of terms belonging to the second group in the vicinity of the pseudo-map graphic in consideration of the positions of the plurality of music artists mapped to the pseudo-map graphic, and moves the display of the term in conjunction with the movement of the pseudo-map graphic.

2. The music artist retrieval system according to claim 1 , wherein the similarity computing section computes a first average value by averaging a plurality of shortest distances measured from each of the plurality of music data for the one music artist to the plurality of music data for the other one music artist, as well as a second average value by averaging a plurality of shortest distances measured from each of the plurality of music data for the other one music artist to the plurality of music data for the one music artist, and then determines a larger one of the first and the second average values as the similarity between the one music artist and the other one music artist.

3. The music artist retrieval system according to claim 1 , wherein the specific artist selecting section includes a rotary selecting section equipped with a rotary knob to be turned by the operator.

4. The music artist retrieval system according to claim 1 , wherein the relevant term display section displays top two or more terms having the term frequency or the term weight which is higher than those of other terms among the plurality of terms belonging to the first group as well as top two or more terms having the term frequency or the term weight which is higher than those of other terms among the plurality of terms belonging to the second group, and increases the brightness of the term, or changes the color of the term, or blinks the term so that the term may be identified as being associated with the one music artist.

5. The music artist retrieval system according to claim 4 , wherein the term data storing section stores term data regarding the term frequency or the term weight for a plurality of terms belonging to a third group, which are considered to be associated with the plurality of music artists; and the relevant term display section displays a term associated with the one music artist as a reference term on the display separately from terms belonging to the first and second groups, based on the term data of the plurality of terms belonging to the third group.

6. The music artist retrieval system according to claim 1 , wherein the closed-loop pseudo-map graphic is constituted from concentrically disposed circle indications of which the number is the same as the number of the top two or more terms belonging to the first group; the top two or more terms corresponds to the two or more concentrically disposed circle indications, respectively; and the pseudo-map graphic display section smoothes the term frequencies or the term weights for the top two or more terms and classifies them into a plurality of levels; constructs the circle indications corresponding to the terms from a plurality of segment regions corresponding to the plurality of levels; and indicates the plurality of levels by using a different brightness and/or chromaticity for each segment region.

7. The music artist retrieval system according to claim 1 , wherein the character information data include data collected via the Internet.